Automated Determination of Neutrophil VCS Parameters in Diagnosis and Treatment Efficacy of Neonatal Sepsis

nature publishing group Clinical Investigation Articles

Automated determination of neutrophil VCS parameters in diagnosis and treatment efficacy of neonatal sepsis

Istemi H. Celik1, Gamze Demirel1, Hatice T. Aksoy1, Omer Erdeve1, Ece Tuncer3, Zeynep Biyikli2 and Ugur Dilmen1

INTRODUCTION: The Coulter LH780 hematology analyzer can The complete blood cell count and peripheral blood smear evaluate mean neutrophil volume (MNV), conductivity (MNC), were the most commonly ordered tests for the diagnosis of sep- scatter (MNS), and distribution width (DW). We sought to investi- sis for many years. Higher white blood cell (WBC) and absolute gate the value of volume, conductivity, and scatter (VCS) param- neutrophil count and, immature/total neutrophil ratio are related eters in diagnosis and treatment efficacy of neonatal sepsis. to sepsis. Peripheral blood smears can give diagnostic informa- RESULTS: We observed significant increases in MNV, volume tion by identifying characteristic morphologic changes seen in distribution width (VDW), conductivity distribution width reactive neutrophils and monocytes (6,7). Size of the cell; den- (CDW), and significant decreases in MNC and MNS in septic sity of the nuclear chromatin; presence of the nucleolus; and the newborns. There were significant decreases in MNV, VDW, and presence of toxic granulation, vacuolization, and Döhle bodies CDW, whereas MNC and MNS increased at the end of the treat- are evaluated to diagnose sepsis. Immature forms of neutrophils ment. Gram-negative sepsis caused higher MNV and VDW than can be identified. However, this approach is labor-intensive and Gram-positive sepsis. time-consuming because it requires manual examination and an DISCUSSION: This is the largest reported study seeking to deter- experienced medical technologist. Furthermore, the results may mine cutoff levels of neutrophil VCS parameters in diagnosis of be subjective because of human interpretation, and only a few sepsis, and the first study in the evaluation of treatment efficacy hundred cells can be analyzed for any given sample (8). and the effects of sepsis onset time and birth weight. We suggest The volume, conductivity, and scatter (VCS) technology of that neutrophil VCS parameters and their DWs are useful both for the Coulter LH 780 hematology analyzer (Beckman Coulter, early diagnosis and evaluation of treatment efficacy in neonatal Fullerton, CA) can obtain data from more than 8,000 WBCs sepsis without requirement for any extra blood collection. using direct current impedance to measure cell volume (V) for METHODS: Peripheral blood samples from 304 newborns, 206 accurate size of all cell types, radio frequency opacity to charac- in group I (76 proven and 130 clinical sepsis) and 98 in group II terize conductivity (C) for internal composition of each cell, and (control group), were studied on diagnosis, 3rd day, and at the a laser beam to measure light scatter (S) for cytoplasmic granu- end of the treatment. larity and nuclear structure. It has been previously reported that VCS parameters can detect morphologic changes in immature eonatal sepsis remains an important clinical syndrome and reactive neutrophils. This so-called VCS technology is anal- Ndespite advances in neonatology. Early diagnosis and ade- ogous to microscopic evaluation of a peripheral blood smear but quate antibiotic treatment are required because of high rates of evaluates more than a microscope (9–13). mortality and morbidity, especially in developing countries (1). VCS parameters were evaluated to diagnose sepsis and mean Early diagnosis of neonatal sepsis is difficult because of non- neutrophil volume (MNV), and volume distribution width specific signs and symptoms, and noninfectious diseases may (VDW) and mean neutrophil scatter (MNS) were found use- mimic neonatal sepsis. Diagnosis is made by clinical and labo- ful in an adult population (11,14). There is only one study in a ratory findings. Blood culture is the gold standard laboratory neonatal population and it showed that MNV and VDW were technique for the diagnosis of infection, but culture results may useful to diagnose neonatal sepsis (13). VCS parameters seem take 48–72 h. False-negative culture results may also occur (2). to be promising indicators for the diagnosis of sepsis. Acute-phase reactants have been used to diagnose sepsis since We sought to (i) investigate the value of VCS parameters in the the 1980s. Several interleukins (ILs), tumor necrosis factor, diagnosis of neonatal sepsis, (ii) evaluate the treatment efficacy C-reactive protein (CRP), procalcitonin, immunoglobulins, and with serial measurements of VCS parameters, and (iii) determine other markers have been used in the diagnosis of sepsis (3,4). the most appropriate cutoff value of each marker in neonatal sep- Cutoff levels of IL-6 and CRP were determined in neonatal sepsis, by using receiver operating characteristic curves and by iden- sis. There is no perfect marker to diagnose sepsis and combina- tifying the diagnostic sensitivity, specificity, positive predictive tion of these markers strengthens the diagnosis (5). value (PPV), and negative predictive value (NPV) of each cutoff.

1Division of Neonatology, Zekai Tahir Burak Maternity Teaching Hospital, Ankara, Turkey; 2Department of Biostatistics, University of Ankara, Ankara, Turkey; 3Blood Bank, Zekai Tahir Burak Maternity Teaching Hospital, Ankara, Turkey. Correspondence: Istemi H. Celik ([email protected])

Received 15 June 2011; accepted 3 August 2011. doi:10.1038/pr.2011.16

Results levels between the sepsis groups and the control group (P < 0.05). We enrolled 304 patients including 76 with proven sepsis, 130 There was no statistical difference between sepsis subgroups for with clinical sepsis, and 98 controls in the study. The charac- all parameters (P > 0.05). teristics of the patients and their distribution within the groups The optimal cutoff levels for MNV, VDW, MNC, CDW, are listed in Table 1. Male/female ratio, gestational age, birth and MNS between the sepsis and control groups and also weight, and vaginal delivery rates were similar between groups between the sepsis subgroups and control group were cal- I and II (P > 0.05). Group Ia had statistically significant lower culated by drawing receiver operating characteristic curves. birth weight and gestational age, and higher rate of late-onset Table 3 shows the cutoff levels. The sepsis group and sepsis sepsis than group Ib (P < 0.05). Twelve types of microorgan- subgroups had similar cutoff levels vs. the control group. The ism, including 43 Gram-positive and 33 Gram-negative bacte- optimal cutoff levels of MNV, VDW, MNC, CDW, and MNS ria, were isolated from blood cultures. The most frequently iso- between groups I and II were >157.15 arbitrary units (au), lated microorganisms were Staphylococcus epidermidis (35.5%), >37.44 au, <159.3 au, >12.3 au, and <127.5 au, respectively. Acinetobacter baumannii (26.3%), Klebsiella pneumoniae The optimal cutoff levels of IL-6 and CRP were 18.9 pg/ml (17.1%), and Escherichia coli (7.9%). and 7.6 mg/dl, respectively, in our sepsis groups. MNV, IL-6, Hemoglobin, WBC, and platelet counts were evaluated. and CRP were found to be independent risk factors of sepsis WBC count was significantly higher in group I (P < 0.05) and in logistic regression analysis. Sensitivity, specificity, PPV, a level above 13,150/μl was calculated as cutoff. The cutoff level and NPV values of MNV, VDW, MNC, CDW, MNS, IL-6, for absolute neutrophil count was 7,650/μl. Platelet count was CRP, and combination of MNV, IL-6, and CRP are listed significantly lower in group I (P < 0.05) and the cutoff level for in Table 4. Sensitivity, specificity, PPV, and NPV of MNV platelet count was below 286,000/μl. Mean hemoglobin levels were 79, 82, 90, and 65%, respectively. The combination of were similar between groups (P > 0.05). MNV, IL-6, and CRP in the diagnosis of sepsis demonstrated We evaluated MNV, mean neutrophil conductivity (MNC), sensitivity, specificity, PPV, and NPV of 94, 88, 95, and 86, MNS, and distribution width (DW) of these parameters. Results respectively. are listed in Table 2. There were significant differences for MNV, Patients with Gram-negative microorganisms had higher VDW, MNC, conductivity distribution width (CDW), and MNS MNV (>163.25 au) and VDW (>41.75 au) levels than patients with Gram-positive microorganisms (P < 0.05). Table 1. Characteristics of patients according to group We also investigated VCS parameters in the evaluation of treatment efficiency with serial measurements. Values of these Group Ia Group Ib Group II (n = 76) (n = 130) (n = 98) Table 3. Cutoff levels of MNV, VDW, MNC, CDW, and MNS Male/female 34/32 67/63 50/48 MNV VDW MNC CDW MNS Gestational age (wk) 30 ± 5 (22–41) 33 ± 4 (24–42) 30 ± 4 (22–40) Group I vs. II >157.15 >37.44 <159.3 >12.3 <127.5 Gestational age 75 52 75 Group Ia vs. II >159.65 >36.92 <159.3 >15.0 <125.15 ≤ 32 wk (%) Group Ib vs. II >157.15 >37.46 <159.3 >12.48 <125.1 Birth weight (g) 1,423 ± 828 1,967 ± 1,035 1,571 ± 626 CDW, conductivity distribution width; MNC, mean neutrophil conductivity; MNS, mean Sepsis work-up day 10 (1–134) 5 (1–80) 11.5 (2–62) neutrophil scatter; MNV, mean neutrophil volume; VDW, volume distribution width. (median)

Late-onset sepsis (%) 89.5 63.5 — Table 4. Sensitivity, specificity, PPV, NPV, and area under the ROC curve for MNV, VDW, MNC, CDW, MNS, IL-6, CRP, and combination of Vaginal delivery (%) 43.4 39.2 38.5 MNV, IL-6, and CRP at the optimal cutoff levels between groups I and II Sensitivity Specificity PPV NPV AUC Cutoff (%) (%) (%) (%) (%) Table 2. Neutrophil volume, conductivity, and scatter parameters MNV >157.15 au 79 82 90 65 85 and distribution width results VDW >37.44 au 60 78 85 48 74 Group I Group Ia Group Ib Group II (n = 206) (n = 76) (n = 130) (n = 98) P MNC <159.3 au 66 64 80 47 66 MNV (au) 170.2 (18.9) 173.4 (22.2) 168.4 (16.6) 148.4 (11.1) <0.05 CDW >12.3 au 78 51 77 52 66 VDW 40.5 (8.5) 42.1 (8.2) 39.6 (8.3) 33.8 (6.0) <0.05 MNS <127.5 au 60 65 21 55 62 MNC (au) 155.9 (11.9) 158.1 (12.9) 154.6 (11.1) 161.5 (9.1) <0.05 IL-6 >18.9 pg/ml 82 93 97 67 91 CDW 15.3 (4.6) 15.7 (5.0) 15.1 (4.4) 13.1 (3.4) <0.05 CRP >7.6 mg/dl 72 99 99 54 89 MNS (au) 125.5 (11.8) 127 (14.8) 125 (9.5) 129 (10.3) <0.05 IL-6, CRP, and MNV 94 88 95 86 94 SDW 17.1 (5.0) 17.4 (5.8) 16.9 (4.5) 17.2 (5.5) >0.05 au, arbitrary units; AUC, area under the curve; CDW, conductivity distribution width; au, arbitrary units; CDW, conductivity distribution width; MNC, mean neutrophil CRP, C-reactive protein; IL, interleukin; MNC, mean neutrophil conductivity; MNS, mean conductivity; MNS, mean neutrophil scatter; MNV, mean neutrophil volume; SDW, neutrophil scatter; MNV, mean neutrophil volume; NPV, negative predictive value; scatter distribution width; VDW, volume distribution width. Values represent mean (SD). PPV, positive predictive value; ROC, receiver operating characteristic; VDW, volume P value is for difference between groups I and II. distribution width.

Table 5. Changes of MNV, VDW, MNC, CDW, and MNS according to >12.3 au, and <127.5 au, respectively. The optimal cutoff levels of serial measurements IL-6 and CRP were 18.9 pg/ml and 7.6 mg/dl, respectively, in our MNV VDW MNC CDW MNS study group. Sensitivity, specificity, PPV, and NPV of MNV and Sepsis work-up day 170.5 41.0 155.5 15.1 125.9 VDW were 79, 82, 90, and 65% and 60, 78, 85, and 48%, respec- Treatment day 3 159.7 37.6 158.0 14.1 122.3 tively. The combination of IL-6, CRP, and MNV in the diagnosis of sepsis gave sensitivity, specificity, PPV, and NPV of 94, 88, 95, At the end of treatment 150.6 36.8 157.3 12.8 129.2 and 86%, respectively. This combination had higher sensitivity CDW, conductivity distribution width; MNC, mean neutrophil conductivity; MNS, mean neutrophil scatter; MNV, mean neutrophil volume; VDW, volume distribution width. and NPV than usage of each parameter alone. Lee et al. (2010) reported that MNC (<145.6 au) and CDW (>10.4 au) obtained parameters from sepsis work-up day to 3rd day and at the end of from cord blood were the predictive parameters for early neona- the treatment are listed in Table 5. There were statistically signifi- tal sepsis in newborns with chorioamnionitis (15). cant decreases in MNV, VDW, and CDW levels with antibiotic Koening et al. (2010) demonstrated higher VDW in bacterial treatment (P < 0.05). MNC and MNS levels increased at the end infections and higher lymphocyte VDW in viral infections in of the treatment but there was no statistical difference (P > 0.05). childhood (16). Chaves et al. (2005) evaluated VCS parameters Sepsis work-up day did not influence the VCS parameters. in 69 adult patients with proven sepsis (11). They found higher Patients with early- and late-onset sepsis had similar VCS levels MNV and lower MNS in the sepsis group in comparison to the (P > 0.05). Patients with birth weight below 1,500 g had higher control group. Cutoff levels of MNV and MNS were suggested VDW and scatter distribution width (SDW) than patients with as 150 au and 141 au, and sensitivity and specificity of MNV and birth weight above 1,500 g (P < 0.05). MNS were 70 and 91%, and 46 and 80%, respectively. Chaves et al. (2006) evaluated VDW and they calculated its cutoff as 22 au Discussion with 79% sensitivity and 94% specificity, and as 23 au with 69% The automated hematology analyzers can rapidly examine sensitivity and 100% specificity, respectively (14). Mardi et al. very large numbers of cells to provide comprehensive hema- (2010) studied neutrophil and monocyte VCS parameters in tology profiles. These parameters can be measured by the adults including 37 sepsis and 39 nonsystemic-infection patients VCS technology of the Coulter LH 780 hematology analyzer (12). They found higher MNV in the sepsis group than in the (9–13). This so-called VCS technology is analogous to micro- nonsystemic-infection and control groups, and lower MNS in scopic evaluation of a peripheral blood smear but evaluates the sepsis and nonsystemic-infection groups than in the con- more than a microscope. This technology gives the results in trol group. They designed cutoff levels of MNV as 150 au with minutes, with no additional blood needed for the complete 76% sensitivity and 63% specificity, and as 145 au with 87% blood cell count. sensitivity and 41% specificity, respectively. They found higher This study, which included 76 septic neonates with posi- mean monocyte volume (MMV) in the sepsis group than in the tive blood cultures, 130 neonates with clinical sepsis, and nonsystemic-infection and control groups, and higher mono- 98 healthy neonates, evaluated neutrophil VCS parameters cyte conductivity in the sepsis and nonsystemic-infection groups in the early diagnosis of sepsis and treatment efficacy, and than in the control group. They designed cutoff levels of MMV sought to establish the cutoff levels, sensitivity, specificity, as 170 au with 86% sensitivity and 40% specificity, and as 175 au PPV, and NPV for each parameter. To our knowledge, this with 60% sensitivity and 54% specificity. is the largest reported study seeking to determine cutoff lev- Patients with Gram-negative microorganisms had higher els of neutrophil VCS parameters in the diagnosis of sepsis MNV (>163.25 au) and VDW (>41.75 au) levels than patients and the first study investigating the neutrophil VCS param- with Gram-positive microorganisms in our study (P < 0.05). eters’ value in evaluating treatment efficacy. This study also This may be associated with high cytokine production potential gives additional information about effects of sepsis onset of Gram-negative microorganisms (17,18). time and birth weight on VCS parameters. Previous stud- We also evaluated the value of these parameters to pre- ies mostly included adult patients and there was only one dict treatment efficacy. MNV, VDW, MNC, CDW, and MNS, study including newborns to diagnose late-onset neonatal which were useful in diagnosing neonatal sepsis, were also sepsis (11–14). found to predict the treatment efficacy with serial measure- Raimondi et al. (2010) reported the cutoff levels of MNV and ments. They all tended to be within normal values during the VDW as 148 au and 27.5 au in late-onset sepsis of very-low- antibiotic treatment. These parameters are evaluated without birth-weight infants, including 20 with proven and 20 with clin- additional specimen requirements. They are quantitative, ical sepsis. The sensitivity, specificity, PPV, and NPV of MNV more objective, and more accurate than manual differential and VDW in this study were 95, 88, 61.3, and 98.9% and 80, counts because more than 8,000 WBCs are evaluated auto- 52, 25, and 92.8%, respectively (13). Combination of MNV and matically (9). CRP gave sensitivity, specificity, PPV, and NPV as 95, 97, 86.4, Sepsis work-up day did not influence the VCS parameters. and 99%, respectively. In addition to this small-size study, we Patients with early- and late-onset sepsis had similar VCS levels. calculated cutoff levels of MNV, VDW, MNC, CDW, and MNS. Patients with birth weight below 1,500 g had higher VDW and The optimal cutoff levels of MNV, VDW, MNC, CDW, and MNS SDW than patients with birth weight above 1,500 g. According between groups I and II were >157.15 au, >37.44 au, <159.3 au, to these data, neutrophil VCS parameters seem not to differ

Copyright © 2012 International Pediatric Research Foundation, Inc. Volume 71 | Number 1 | January 2012 Pediatric Research 123 Articles Celik et al. within sepsis day and birth weight and can be used to diagnose pass through the helium-neon laser beam, provide information about neonatal sepsis in all newborns. cell surface characteristics and cell granularity. The raw data events col- Neonatal sepsis remains one of the most important clinical lected from volume, conductivity, and light scatter are plotted three- dimensionally on a data plot. All these measurements (mean and dis- syndromes, especially in preterm infants, leading to high mor- tribution width for volume, conductivity, and scatter) are reported as bidity and mortality, despite advances in neonatology. Although numerical values (cell population data) for each sample. early diagnosis and treatment is needed, because of nonspecifc VCS parameters were evaluated on the same day of sepsis diagnosis, signs and symptoms, difficulties and limitations of culturing pro- 3rd day, and at the end of the treatment. cedures, and the lack of a perfect indicator of sepsis, diagnosis can Statistical Analysis be delayed. We report cut-off levels of neutrophil VCS parameters Statistical analyses were performed using the SPSS statistical package in the diagnosis of sepsis, and these parameters seem to be useful (v. 15.0). Categorical variables between groups were analyzed using the in evaluating indicators of treatment efficacy such as CRP. χ2 test. Comparison of means between two groups was examined by In conclusion, we suggest that neutrophil VCS parameters and using a t test, where the data fit a normal distribution, and by Mann- Whitney U test, where the data were nonnormal distributions. For their DWs are useful both for early diagnosis and evaluation of comparison of more than two groups, ANOVA was used for normal treatment efficacy in neonatal sepsis without requirement for distributions and the Kruskal-Wallis test for nonnormal distributions. any extra blood collection. Receiver operating characteristic curve analysis was used to determine the power of variables to differentiate groups, and the area under the Methods curve was calculated; significant cutoff levels were calculated using a Settings Youden index. Logistic regression analysis was used to compare independent variables and then sensitivity and specificity were calculated This study was conducted at the neonatal intensive care unit of Zekai for independent variables determined by logistic regression analysis. A Tahir Burak Maternity Teaching Hospital, the biggest tertiary neo- P value of <0.05 was deemed to indicate statistical significance. natology unit in Turkey, between October 2010 and April 2011. Our hospital’s birth, neonatal intensive care unit admission, and patient day numbers are approximately 20,000, 4,000, and 47,000 per year, respec- Ethics tively. Percentages of patients with sepsis and proven sepsis are 9.2% This study was approved by the local ethics committee of Zekai Tahir and 3.4%, respectively. Burak Maternity Teaching Hospital. An informed consent was obtained from all patients’ parents.

Participants and Definitions Disclosure: This study was accepted and presented as a poster presentation Clinical findings for a diagnosis of sepsis required at least three of the fol- at the 24th International Society of Laboratory Hematology (ISLH) Congress, lowing: bradycardia (<100/min), tachycardia (>200/min), hypotension, New Orleans, LA. hypotonia, seizures, apnea, tachypnea, cyanosis, respiratory distress, poor skin color and perfusion, feeding difficulty, irritability, lethargy, REFERENCES and laboratory results showing elevated levels of IL-6 and/or CRP (19). 1. Osrin D, Vergnano S, Costello A. Serious bacterial infections in newborn Patients were grouped into a sepsis group (group I), including infants in developing countries. Curr Opin Infect Dis 2004;17:217–24. proven and clinical sepsis, and a control group (group II): Group Ia 2. Gerdes JS. Diagnosis and management of bacterial infections in the neo- (proven sepsis): newborns with positive blood cultures, clinical find- nate. Pediatr Clin North Am 2004;51:939–59, viii–ix. ings in agreement with the diagnosis, and elevated IL-6 and/or CRP 3. Khassawneh M, Hayajneh WA, Kofahi H, Khader Y, Amarin Z, Daoud levels during the clinical course; Group Ib (clinical sepsis): newborns A. Diagnostic markers for neonatal sepsis: comparing C-reactive protein, with clinical findings of infection, plus a significant increase in IL-6 interleukin-6 and immunoglobulin M. Scand J Immunol 2007;65:171–5. and/or CRP levels during the clinical course, but with negative blood 4. Dilli D, Oguz SS, Dilmen U, Köker MY, Kizilgün M. Predictive val- cultures; Group II (control group): newborns admitted to the hos- ues of neutrophil CD64 expression compared with interleukin-6 and pital for perinatal risk factors such as prematurity; ablatio placenta; C-reactive protein in early diagnosis of neonatal sepsis. J Clin Lab Anal Rh isoimmunization; conditions, such as hypoglycemia; intrauterine 2010;24:363–70. growth restriction; transient tachypnea; and indirect hyperbilirubine- 5. Celik IH, Demirel FG, Uras N, et al. What are the cut-off levels for IL-6 and mia without clinical findings of infection were used to define control CRP in neonatal sepsis? J Clin Lab Anal 2010;24:407–12. levels. Infants in the control group had normal physical examination 6. Mathy KA, Koepke JA. The clinical usefulness of segmented vs. stab findings and were matched as far as possible in demographic charac- neutrophil criteria for differential leukocyte counts. Am J Clin Pathol teristics to those in the proven and clinical sepsis groups. 1974;61:947–58. 7. Wenz B, Gennis P, Canova C, Burns ER. The clinical utility of the leuko- Techniques cyte differential in emergency medicine. Am J Clin Pathol 1986;86:298– Before antibiotic therapy, blood samples for VCS parameters and cul- 303. ture were taken from patients with a diagnosis of sepsis. 8. Pierre RV. Peripheral blood film review. The demise of the eyecount leukocyte differential. Clin Lab Med 2002;22:279–97. Data Collection 9. Richardson-Jones A. An automated hematology instrument for compre- Data included total WBC counts, percentage of neutrophils, and the hensive WBC, RBC, and platelet analysis. Am Clin Lab 1990;9:18–22. VCS parameters of neutrophils, which were generated by each individ- 10. Krause JR. Automated differentials in the hematology laboratory. Am J ual cell passing through the aperture and were optically and electroni- Clin Pathol 1990;93(4 Suppl 1):S11–6. cally measured by the Coulter LH 780 (Beckman Coulter, Fullerton, 11. Chaves F, Tierno B, Xu D. Quantitative determination of neutrophil VCS CA). The reagents that are used in the LH780 to prepare the sample for parameters by the Coulter automated hematology analyzer: new and reliable the analysis allow preserving WBCs in near-native state and remov- indicators for acute bacterial infection. Am J Clin Pathol 2005;124:440–4. ing all red blood cells with the lytic solution. Volume, as measured by 12. Mardi D, Fwity B, Lobmann R, Ambrosch A. Mean cell volume of neutro- direct current, is used to identify the size of the cell. Conductivity, or phils and monocytes compared with C-reactive protein, interleukin-6 and radio frequency measurement, provides information about the internal white blood cell count for prediction of sepsis and nonsystemic bacterial characteristics of the cell. Light-scatter measurements, obtained as cells infections. Int J Lab Hematol 2010;32:410–8.

13. Raimondi F, Ferrara T, Capasso L, et al. Automated determination of 16. Koenig S, Quillen K. Using neutrophil and lymphocyte VCS indices in neutrophil volume as screening test for late-onset sepsis in very low birth ambulatory pediatric patients presenting with fever. Int J Lab Hematol infants. Pediatr Infect Dis J 2010;29:288. 2010;32:449–51. 14. Chaves F, Tierno B, Xu D. Neutrophil volume distribution width: a new 17. Abe R, Oda S, Sadahiro T, et al. Gram-negative bacteremia induces greater automated hematologic parameter for acute infection. Arch Pathol Lab magnitude of inflammatory response than Gram-positive bacteremia. Crit Med 2006;130:378–80. Care 2010;14:R27. 15. Lee JC, Ahern TP, Chaves FP, Quillen K. Utility of hematologic 18. Webb SA, Kahler CM. Bench-to-bedside review: bacterial virulence and and volume, conductivity, and scatter parameters from umbilical subversion of host defences. Crit Care 2008;12:234. cord blood in predicting chorioamnionitis. Int J Lab Hematol 2010; 19. Haque KN. Definitions of bloodstream infection in the newborn. Pediatr 32:351–9. Crit Care Med 2005;6(3 Suppl):S45–9.